btrfs-progs/file.c

331 lines
8.2 KiB
C

/*
* Copyright (C) 2014 Fujitsu. All rights reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public
* License v2 as published by the Free Software Foundation.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* General Public License for more details.
*
* You should have received a copy of the GNU General Public
* License along with this program; if not, write to the
* Free Software Foundation, Inc., 59 Temple Place - Suite 330,
* Boston, MA 021110-1307, USA.
*/
#include <sys/stat.h>
#include "ctree.h"
#include "utils.h"
#include "disk-io.h"
#include "transaction.h"
#include "kerncompat.h"
/*
* Get the first file extent that covers (part of) the given range
* Unlike kernel using extent_map to handle hole even no-hole is enabled,
* progs don't have such infrastructure, so caller should do extra care
* for no-hole.
*
* return 0 for found, and path points to the file extent.
* return >0 for not found, and path points to the insert position.
* return <0 for error.
*/
int btrfs_get_extent(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
struct btrfs_path *path,
u64 ino, u64 offset, u64 len, int ins_len)
{
struct btrfs_key key;
struct btrfs_key found_key;
struct btrfs_file_extent_item *fi_item;
u64 end = 0;
int ret = 0;
int not_found = 1;
key.objectid = ino;
key.type = BTRFS_EXTENT_DATA_KEY;
key.offset = offset;
ret = btrfs_search_slot(trans, root, &key, path, ins_len,
ins_len ? 1 : 0);
if (ret <= 0)
goto out;
if (ret > 0) {
/* Check previous file extent */
ret = btrfs_previous_item(root, path, ino,
BTRFS_EXTENT_DATA_KEY);
if (ret < 0)
goto out;
if (ret > 0)
goto check_next;
}
btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
if (found_key.objectid != ino ||
found_key.type != BTRFS_EXTENT_DATA_KEY)
goto check_next;
fi_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
struct btrfs_file_extent_item);
end = found_key.offset +
btrfs_file_extent_ram_bytes(path->nodes[0], fi_item);
/*
* existing file extent
* |--------| |----|
* |-------|
* offset + len
* OR
* |---------------|
* |-------|
*/
if (end > offset) {
not_found = 0;
goto out;
}
check_next:
ret = btrfs_next_item(root, path);
if (ret)
goto out;
btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
if (found_key.objectid != ino ||
found_key.type != BTRFS_EXTENT_DATA_KEY) {
ret = 1;
goto out;
}
if (found_key.offset < offset + len)
/*
* existing file extent
* |---| |------|
* |-------|
* offset + len
*/
not_found = 0;
else
/*
* existing file extent
* |----| |----|
* |----|
* offset + len
*/
not_found = 1;
/*
* To keep the search behavior consistent with search_slot(),
* we need to go back to the prev leaf's nritem slot if
* we are at the first slot of the leaf.
*/
if (path->slots[0] == 0) {
ret = btrfs_prev_leaf(root, path);
/* Not possible */
if (ret)
goto out;
path->slots[0] = btrfs_header_nritems(path->nodes[0]);
}
out:
if (ret == 0)
ret = not_found;
return ret;
}
/*
* Punch hole ranged [offset,len) for the file given by ino and root.
*
* Unlink kernel punch_hole, which will not zero/free existing extent,
* instead it will return -EEXIST if there is any extents in the hole
* range.
*/
int btrfs_punch_hole(struct btrfs_trans_handle *trans,
struct btrfs_root *root,
u64 ino, u64 offset, u64 len)
{
struct btrfs_path *path;
int ret = 0;
path = btrfs_alloc_path();
if (!path)
return -ENOMEM;
ret = btrfs_get_extent(NULL, root, path, ino, offset, len, 0);
if (ret < 0)
goto out;
if (ret == 0) {
ret = -EEXIST;
goto out;
}
ret = btrfs_insert_file_extent(trans, root, ino, offset, 0, 0, len);
out:
btrfs_free_path(path);
return ret;
}
/*
* Read out content of one inode.
*
* @root: fs/subvolume root containing the inode
* @ino: inode number
* @start: offset inside the file, aligned to sectorsize
* @len: length to read, aligned to sectorisize
* @dest: where data will be stored
*
* NOTE:
* 1) compression data is not supported yet
* 2) @start and @len must be aligned to sectorsize
* 3) data read out is also aligned to sectorsize, not truncated to inode size
*
* Return < 0 for fatal error during read.
* Otherwise return the number of succesfully read data in bytes.
*/
int btrfs_read_file(struct btrfs_root *root, u64 ino, u64 start, int len,
char *dest)
{
struct btrfs_key key;
struct btrfs_path path;
struct extent_buffer *leaf;
struct btrfs_inode_item *ii;
u64 isize;
int no_holes = btrfs_fs_incompat(root->fs_info, NO_HOLES);
int slot;
int read = 0;
int ret;
if (!IS_ALIGNED(start, root->fs_info->sectorsize) ||
!IS_ALIGNED(len, root->fs_info->sectorsize)) {
warning("@start and @len must be aligned to %u for function %s",
root->fs_info->sectorsize, __func__);
return -EINVAL;
}
btrfs_init_path(&path);
key.objectid = ino;
key.offset = start;
key.type = BTRFS_EXTENT_DATA_KEY;
ret = btrfs_search_slot(NULL, root, &key, &path, 0, 0);
if (ret < 0)
goto out;
if (ret > 0) {
ret = btrfs_previous_item(root, &path, ino, BTRFS_EXTENT_DATA_KEY);
if (ret > 0) {
ret = -ENOENT;
goto out;
}
}
/*
* Reset @dest to all 0, so we don't need to care about holes in
* no_hole mode, but focus on reading non-hole part.
*/
memset(dest, 0, len);
while (1) {
struct btrfs_file_extent_item *fi;
u64 extent_start;
u64 extent_len;
u64 read_start;
u64 read_len;
u64 read_len_ret;
u64 disk_bytenr;
leaf = path.nodes[0];
slot = path.slots[0];
btrfs_item_key_to_cpu(leaf, &key, slot);
if (key.objectid > ino)
break;
if (key.type != BTRFS_EXTENT_DATA_KEY || key.objectid != ino)
goto next;
extent_start = key.offset;
if (extent_start >= start + len)
break;
fi = btrfs_item_ptr(leaf, slot, struct btrfs_file_extent_item);
if (btrfs_file_extent_compression(leaf, fi) !=
BTRFS_COMPRESS_NONE) {
ret = -ENOTTY;
break;
}
/* Inline extent, one inode should only one inline extent */
if (btrfs_file_extent_type(leaf, fi) ==
BTRFS_FILE_EXTENT_INLINE) {
extent_len = btrfs_file_extent_inline_len(leaf, slot,
fi);
if (extent_start + extent_len <= start)
goto next;
read_extent_buffer(leaf, dest,
btrfs_file_extent_inline_start(fi), extent_len);
read += round_up(extent_len, root->fs_info->sectorsize);
break;
}
extent_len = btrfs_file_extent_num_bytes(leaf, fi);
if (extent_start + extent_len <= start)
goto next;
read_start = max(start, extent_start);
read_len = min(start + len, extent_start + extent_len) -
read_start;
/* We have already zeroed @dest, nothing to do */
if (btrfs_file_extent_type(leaf, fi) ==
BTRFS_FILE_EXTENT_PREALLOC ||
btrfs_file_extent_disk_num_bytes(leaf, fi) == 0) {
read += read_len;
goto next;
}
disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi) +
btrfs_file_extent_offset(leaf, fi);
read_len_ret = read_len;
ret = read_extent_data(root, dest + read_start - start, disk_bytenr,
&read_len_ret, 0);
if (ret < 0)
break;
/* Short read, something went wrong */
if (read_len_ret != read_len)
return -EIO;
read += read_len;
next:
ret = btrfs_next_item(root, &path);
if (ret > 0) {
ret = 0;
break;
}
}
/*
* Special trick for no_holes, since for no_holes we don't have good
* method to account skipped and tailling holes, we used
* min(inode size, len) as return value
*/
if (no_holes) {
btrfs_release_path(&path);
key.objectid = ino;
key.offset = 0;
key.type = BTRFS_INODE_ITEM_KEY;
ret = btrfs_lookup_inode(NULL, root, &path, &key, 0);
if (ret < 0)
goto out;
if (ret > 0) {
ret = -ENOENT;
goto out;
}
ii = btrfs_item_ptr(path.nodes[0], path.slots[0],
struct btrfs_inode_item);
isize = round_up(btrfs_inode_size(path.nodes[0], ii),
root->fs_info->sectorsize);
read = min_t(u64, isize - start, len);
}
out:
btrfs_release_path(&path);
if (!ret)
ret = read;
return ret;
}